Biofilm Growth and Near-Infrared Radiation-Driven Photosynthesis of the Chlorophyll d-Containing Cyanobacterium Acaryochloris marina

Abstract

The cyanobacterium Acaryochloris marina is the only known phototroph harboring chlorophyll (Chl) d. It is easy to cultivate it in a planktonic growth mode, and A. marina cultures have been subject to detailed biochemical and biophysical characterization. In natural situations, A. marina is mainly found associated with surfaces, but this growth mode has not been studied yet. Here, we show that the A. marina type strain MBIC11017 inoculated into alginate beads forms dense biofilm-like cell clusters, as in natural A. marina biofilms, characterized by strong O₂ concentration gradients that change with irradiance. Biofilm growth under both visible radiation (VIS, 400 to 700 nm) and near-infrared radiation (NIR, ∼700 to 730 nm) yielded maximal cell-specific growth rates of 0.38 per day and 0.64 per day, respectively. The population doubling times were 1.09 and 1.82 days for NIR and visible light, respectively. The photosynthesis versus irradiance curves showed saturation at a photon irradiance of E_k (saturating irradiance) >250 μmol photons m⁻² s⁻¹ for blue light but no clear saturation at 365 μmol photons m⁻² s⁻¹ for NIR. The maximal gross photosynthesis rates in the aggregates were ∼1,272 μmol O₂ mg Chl d⁻¹ h⁻¹ (NIR) and ∼1,128 μmol O₂ mg Chl d⁻¹ h⁻¹ (VIS). The photosynthetic efficiency (α) values were higher in NIR-irradiated cells [(268 ± 0.29) × 10⁻⁶ m² mg Chl d⁻¹ (mean ± standard deviation)] than under blue light [(231 ± 0.22) × 10⁻⁶ m² mg Chl d⁻¹]. A. marina is well adapted to a biofilm growth mode under both visible and NIR irradiance and under O₂ conditions ranging from anoxia to hyperoxia, explaining its presence in natural niches with similar environmental conditions.

INTRODUCTION

The cyanobacterium Acaryochloris marina is unique among phototrophs because it has largely replaced chlorophyll a (Chl a) in its photosystems with Chl d. It is reasonably easy to cultivate A. marina as suspended (planktonic) cultures in minimal or enriched medium (31, 32). Due to its unique possession of Chl d, such A. marina cultures have been the subject of a number of detailed studies, including studies of its genome (32, 47), its basic photophysiology (2, 8, 49, 50), Chl d biosynthesis (45), chromatic photoacclimation (18), pigment composition (1, 9), and ultrastructure (28, 48). A recent study also demonstrated the presence and activity of N₂ fixation under hypoxic conditions in a culture of the A. marina strain HICR111A (36).

In contrast, A. marina's ecology and natural habitats are much less studied, but it is mostly reported from surface-associated habitats and has been shown to form aggregates of cells in biofilms. Examples of such A. marina-containing biofilms are the underside of the tropical ascidian Lissoclinum patella (22) and endolithic microbial communities (6) depleted of visible radiation (VIS, 400 to 700 nm) and enriched in near-infrared radiation (NIR, >700 nm). While naturally occurring cells of A. marina apparently prefer to grow associated with surfaces, laboratory investigations of the growth and NIR-dependent photosynthesis of A. marina strains in biofilms are lacking.

While planktonic growth occurs under more homogeneous and less dynamic conditions, microbes in photosynthetic biofilms experience steep gradients in O₂, pH, nutrient availability, light, and temperature due to a high volume-specific metabolic activity coupled with mass transfer limitation by diffusive transport in the dense matrix of cells and exopolymers (23). Changes in irradiance or substrate supply lead to fast dynamic shifts in the physicochemical conditions in biofilms over short time intervals (seconds to minutes). Protection against environmental stress, greater nutrient availability, and acquisition of new genetic traits through horizontal gene transfer are some of the suggested advantages of growing in a biofilm (13), yet many of these proposed benefits remain untested.

Marked differences between planktonic and biofilm growth modes have been observed in bacterial pathogens (15, 16, 38, 44, 53), green algae (19, 43), and yeast (54). In these organisms, the difference in growth forms is reflected in their physiology, gene expression, phenotypic plasticity, and functional characteristics. Therefore, extrapolation of data retrieved from planktonic cultures to natural systems can be difficult (12, 13), highlighting the need to conduct ecologically relevant experiments in the appropriate growth mode.

Biofilm-like growth conditions can be mimicked by incorporating cells into a matrix, such as alginate (52). This forces the entrapped cells to grow in close proximity to each other in a mass transfer-limited matrix, effectively creating microenvironmental conditions similar to those in naturally occurring biofilms. Accordingly, the employment of such immobilization may be of assistance to establish reproducible experimental systems that provide metabolic measures under closer approximation to the biofilm growth mode. In this study, we employed such alginate immobilization to study the biofilm-like growth and photosynthetic activity of the Chl d-containing cyanobacterium A. marina. Cell growth was monitored using flow cytometric cell counts, chlorophyll content, and imaging by microscopy. Simultaneous measurements of gross photosynthesis and effective quantum yields were performed to gain information on the photosynthetic competence and efficiency of immobilized A. marina.

MATERIALS AND METHODS

Acaryochloris marina growth conditions.

Acaryochloris marina strain MBIC11017 was grown in 200-ml cell culture flasks with KESM medium (salinity of 30) on a shaking incubator at 28°C as previously reported (30). The cultures were shaken at 100 rpm under a 12-/12-h light-dark period. Near-infrared radiation (NIR) was provided by narrow-band light-emitting diodes (LEDs) (centered at 720 nm) (Epitex, Inc., Japan) at an irradiance of 20 to 40 μmol photons m⁻² s⁻¹. Another set of cultures were grown under the same irradiance but using visible light delivered by a halogen lamp (HQ Power; Brinck Electronic, Denmark). Planktonic and immobilized A. marina cells were grown under the same light and temperature conditions.

Alginate bead preparation and sampling.

Beads were prepared using alginate extracted from the brown alga Laminaria hyperborea (Protanal LF 10/60; FMC Biopolymer, Drammen, Norway) using a wt/vol ratio of 4.9%. For 100 ml of alginate solution, 4.9 g of Protanal LF 10/60 was suspended in 70 ml of MilliQ water and stirred on a hotplate set to 50°C until dissolved, after which the remaining 30 ml of water was added. The alginate solution was subsequently autoclaved at 120°C for 20 min.

For bead preparation, 15 ml of the alginate solution was mixed with 5 ml of pregrown planktonic Acaryochloris marina culture exhibiting an optical density of 0.5 at 750 nm. Additionally, cells were counted using a Thoma-hemocytometer (Blaubrand GmbH, Germany) and diluted to a concentration of 1.0 × 10⁷ cells ml⁻¹ alginate solution. The alginate-culture solution was vortexed for ∼45 s at the highest speed and then transferred into a sterile 20-ml syringe. The syringe was placed into a syringe pump (Aladdin 220; World Precision Instruments, Sarasota, FL) and connected to a 0.5-mm-wide and 60-mm-long sterile hypodermic needle using sterile tubing. The injection needle was placed 3 cm above a sterilized 250 ml beaker filled with 150 ml of prewarmed (28°C) 4% SrCl₂ solution, previously reported to yield beads of the necessary stability for long-term incubation in high-salinity seawater (33). The solution was stirred at 150 rpm, and the syringe pump was set at a delivery rate of 1 ml min⁻¹, corresponding to a formation of ∼1 bead s⁻¹, with the spherical beads having a final diameter of 2.52 ± 0.17 mm (mean ± standard deviation). The beads were hardened in the SrCl₂ solution for 45 min and were then washed three times with deionized water (Maxima; Holm & Halby, Denmark) to remove any remaining SrCl₂. After washing, the beads were placed into prewarmed KESM medium (28°C) and grown under the same conditions as planktonic cultures with the exception that the medium was changed every 4th day.

Beads were sampled in technical triplicates every 96 h by use of a sterile plastic syringe, placed into microcentrifuge tubes, and immediately transferred into a −80°C freezer for later chlorophyll extraction and flow cytometric analysis.

Chlorophyll extraction and spectrophotometry.

Aliquots of 20 frozen beads were dissolved in 0.5 M Tri-Na-citrate (pH 6.5) for 20 min and homogenized with a sterile micropestle. Cells were centrifuged at 5,000 × g for 10 min, the supernatant removed, and the resulting pellet resuspended in 96% ethanol and incubated at 4°C for 30 min. During incubation, the samples were resuspended with a sterile micropestle every 15 min. The alginate was pelleted by centrifugation at 5,000 × g, and the supernatant was used to determine Chl d and Chl a concentrations spectrophotometrically according to the method in reference 41, using a scanning spectrophotometer (UV-2101PC; Shimadzu, Japan). The chlorophyll content of the beads was calculated in mg cm⁻³, assuming a perfect sphere and even distribution of chlorophyll across individual beads. To prevent bleaching of the photopigments, all handling was done as quickly as possible and under low irradiance.

Flow cytometry.

Aliquots of 20 frozen beads were dissolved in a manner similar to that described for chlorophyll extraction, with the addition of a 30-min sonication step (Bransonic 1510 sonicator; Danbury, CT) after treatment with 0.5 M Tri-Na-citrate (pH 6.5). Following sonication, the cells were filtered through a nylon syringe filter with a mesh size of 10 μm (Frisenette ApS, Denmark) in order to remove any residual alginate. All flow cytometric analyses were performed on a FACSAria III (BD Biosciences, United States), using a 488-nm laser in conjunction with the BD FACSFlow sheath fluid (BD Biosciences, United States). Voltages were set at 140 V for forward scatter (FSC) and 360 V for side scatter (SSC). All parameters were expressed on a logarithmic scale. Flow rates (μl min⁻¹) were determined as previously described (27), and the average flow rate was calculated based on three replicates: before, during, and after enumeration of cells in the samples. Samples were mixed by vortexing, and all gated events were counted for 1 min at flow rate 1, where gates were established based upon FSC and SSC population characteristics. The collected data were analyzed using the software package BD FACSDiva (BD Biosciences, United States). Cell-specific growth rates (μ) and corresponding population doubling times were calculated using the linear increase in cell abundance as determined by semi-log plotting.

Light calibration.

Absolute irradiance measurements of NIR from the LED light sources were performed using a calibrated spectrometer (Jaz; Ocean Optics, Dunedin, FL), where photon irradiance was integrated over a spectral range from 650 to 800 nm. Different NIR irradiances were administered by altering the distance of three collimated NIR LEDs (700 to 740 nm centered at 730 nm) (Roithner Lasertechnik, Vienna, Austria) to the sample. A pulse-amplitude-modulated (PAM) fluorometer (JuniorPAM, Walz GmbH, Germany) was used to administer defined levels of blue actinic light via an optical fiber (1.5 mm diameter); irradiance levels were calibrated with a mini-quantum scalar irradiance sensor (Walz GmbH, Effeltrich, Germany) connected to a quantum irradiance meter (LiCor-250; LiCor, Inc., Lincoln, NE). Ambient light was excluded by covering the experimental setup with black cloth.

Microsensor measurements.

Single beads were immobilized on a piece of black neoprene with thin stainless steel insect preparation needles and submerged in a flow chamber flushed with preheated artificial seawater (28°C, salinity of 30, Marine Environment; Aquacraft, San Carlos, CA). The water was circulated through the flow chamber from an aerated reservoir by means of a submerged water pump.

A Clark-type O₂ microsensor (tip size, ∼20 μm [39]) was mounted on a motorized micromanipulator (Pyro-Science GmbH, Aachen, Germany) and connected to a picoammeter (Unisense PA2000; Unisense A/S, Aarhus, Denmark). The sensor was linearly calibrated from measurements in O₂-free seawater (by the addition of sodium dithionite) and fully aerated seawater at the experimental temperature and salinity. As the alginate beads were flexible and easily compressed, the beads were first carefully punctured by inserting the microsensor to a depth of approximately 1.5 mm; upon retraction, the elastic alginate matrix sealed the hole. For all subsequent microprofiling, the sensor was then inserted into the same position. The position where the microsensor tip touched the bead surface was determined by visual inspection using a stereomicroscope mounted on a stand in front of the aquarium. All microsensor signals were recorded on a strip chart recorder (Kipp and Zonen, B.V., Delft, Netherlands) and via an analog-to-digital (A/D) converter (AD-216; Unisense A/S, Aarhus, Denmark) connected to a personal computer (PC) running data acquisition and micromanipulator-positioning software (Profix; Pyro-Science GmbH, Aachen, Germany).

Variable chlorophyll fluorescence measurements.

A pulse amplitude-modulated (PAM) fluorometer (Junior-PAM; Walz GmbH, Effeltrich, Germany), placed with the 1.5-mm-wide measuring fiber situated close to the bead surface, was used to investigate the photosynthetic activity of A. marina beads via variable chlorophyll fluorescence analysis. A detailed description of such analysis and the measuring principles is beyond the scope of this report and has been described elsewhere (5, 37, 46). Using the saturation pulse method (5, 46), the maximum quantum yield of photosystem II [(ΦPSII)_max] photosynthetic energy conversion, (Φ_PSII)_max = (F_m − F₀)/F_m, was measured after dark incubation and the light-adapted effective quantum yield of PSII, Φ_PSII = (F′_m − F)/F′_m, was measured at defined levels of actinic irradiance. In these equations, F₀ and F denote the minimal fluorescence yields of PSII in darkness and under actinic light illumination, respectively, while F_m and F′_m denote the maximal fluorescence yields of PSII as measured under a strong saturating light pulse in darkness and under actinic light illumination, respectively. The PSII-driven relative electron transport rates (rETR) were calculated with the formula rETR = Φ_PSII × PAR (photosynthetically active radiation). To investigate light acclimatization, rETR-versus-quantum irradiance curves were determined in beads previously incubated under defined levels of either NIR or blue light. Actinic blue light was provided through the Junior-PAM fiber, whereas calibrated levels of NIR were provided by LEDs.

Combined chlorophyll fluorescence and O₂ evolution measurements.

For simultaneous measurements of O₂ concentration and variable chlorophyll fluorescence, the beads were mounted in the flow chamber as described above. The Junior-PAM measuring fiber was situated close to the bead surface and was used to measure variable chlorophyll fluorescence and to provide blue actinic light. Gross photosynthesis (in units of nmol O₂ cm⁻³ s⁻¹) was measured using the O₂ microsensor light-dark shift method (23, 40) with the tip at a suitable distance within the bead. Gross photosynthesis measurements were subsequently normalized to volume-specific Chl d contents found in single beads and calculated into units of μmol O₂ mg Chl d⁻¹ h⁻¹. Consecutive to the light-dark shift, variable chlorophyll fluorescence was determined by applying three saturation pulses for 0.8 s at 10-s intervals with an irradiance of ∼8,000 μmol photons m⁻² s⁻¹. Besides measurements with blue actinic light, similar measurements were also done with defined levels of NIR light delivered from LEDs. More details on variable chlorophyll fluorescence measurements in combination with O₂ microsensor measurements of gross photosynthesis can be found in reference 51.

Light microscopy.

Beads were cut in half using a razorblade and mounted on microscopy slides with the cut surface facing down. Imaging was performed with a coolSNAP camera (Photometrics, Tucson, AZ) mounted on a Zeiss Axioscope 2 plus microscope (Zeiss GmbH, Jena, Germany), using either ×10 or ×40 magnification. Images were recorded at the same time as sampling for flow cytometric analysis, and chlorophyll extraction was performed (approximately every 96 h).

SEM.

Comparative scanning electron microscopy (SEM) analysis was performed on bead cross sections and A. marina-containing biofilm samples from the underside of the didemnid ascidian Lissoclinum patella (see details of sampling in reference 7). Following a rinse in distilled water, cross sections of beads and biofilms isolated from Lissoclinum patella were first dehydrated with 100% ethanol according to standard procedures and were then subjected to critical-point drying (Balzers CPD 030; Balzers, Switzerland) using CO₂. The beads were subsequently mounted on stubs using colloidal coal as an adhesive and sputter coated with gold (SEM coating unit E5000; Polaron). Beads were examined with an XL FEG 30 (Philips, Netherlands) scanning electron microscope operated at an accelerating voltage of 2 kV.

RESULTS

Artificial and natural biofilms of A. marina show similar characteristics.

Beads inoculated with A. marina became visibly green due to growth and accumulation of chlorophyll over a 5-week incubation period (Fig. 1A). Light microscopy of cross-sectioned beads demonstrated a typical growth pattern for immobilized cells: growth started with a few dividing cyanobacteria, the matrix expanded in order to accommodate the additional cells, dense clusters of cyanobacteria were formed, and the expansion of the clusters finally reached the outer perimeter of the alginate matrix, where A. marina cells started to leak into the surrounding medium (Fig. 1B, days 0 to 32). Growth of other, unidentified bacteria was observed in immobilized cells grown under visible light but not in those grown under NIR. Visual inspection of A. marina cell clusters within the alginate matrix showed no further increase in their size after 32 days in both light treatments. Growing the cells for an additional 14 days (data not shown) revealed no additional increase in cell cluster size.

Fig 1.

Visual representation of Acaryochloris marina growth in alginate beads. Beads were grown under either near-infrared radiation (NIR) or visible light (VIS). (A) Macroscopic imaging of A. marina growing in alginate over 5 weeks. (B) Light micrographs of bead cross sections. Images of cross sections were recorded using ×10 or ×40 magnification. Beads were sampled every 4th day with the exception of days 28 and 30, when only NIR (day 28) or VIS beads (day 30) were sampled.

The SEM investigation of immobilized cells in the alginate beads showed cells of approximately 1 to 2 μm in size growing in dense clusters within the alginate matrix (Fig. 2A, C, and E). Growth of other bacteria was not noticed in beads kept under NIR (Fig. 2A and C) but was frequently observed in beads grown under visible light (Fig. 2E). SEM imaging of the underside of the didemnid ascidian Lissoclinum patella, a natural habitat of A. marina (7), showed cells of approximately the same size as the cultured A. marina cells (1 to 2 μm) (Fig. 2B, D, and F). Clusters of A. marina-like cells were frequently found in places where the ascidian test material folded or where other three-dimensional structures occurred.

Fig 2.

Scanning electron micrographs of cross sections through alginate beads with immobilized A. marina cells (A, C, and E) and naturally occurring biofilms from the underside of the didemnid ascidian Lissoclinum patella (B, D, and F). (A, C) Typical cell clusters of A. marina growing within the alginate matrix; note the laminar layers of alginate enclosing the cell clusters. (E) Growth of unidentified bacteria in alginate beads kept under visible light. (B, D, F) naturally occurring biofilms found on the underside of the didemnid ascidian L. patella; note the round A. marina-like cells (1 to 2 μm in size) that closely resemble A. marina cells growing in alginate (see A, C, and E).

Wavelength-dependent photopigment production and cell growth.

Alginate-embedded A. marina cells grown under NIR exhibited approximately 2-fold higher Chl d concentrations than cells grown under visible light (Fig. 3A). Maximal Chl d concentration was observed after 39 days in beads grown under visible light (2.28 ± 0.8 μg Chl d bead⁻¹; n = 3) and 48 days with NIR (5.88 ± 0.4 μg Chl d bead⁻¹; n = 3). The concentration of Chl a was not affected by the growth incubation wavelength (visible light, 0.19 ± 0.08 μg Chl a bead⁻¹, n = 3, and NIR, 0.19 ± 0.01 μg Chl a bead⁻¹, n = 3, at day 40) (Fig. 3B).

Fig 3.

Pigment and cell concentrations of alginate-embedded A. marina. VIS, cells grown under visible light; NIR, cells grown under near-infrared LEDs. All data shown are based on technical triplicates; error bars denote the standard deviation from the mean. (A) Chlorophyll d concentration of cells grown under visible light and NIR. (B) Chlorophyll a concentration of cells grown under NIR and visible light. (C) Cell counts per single bead as determined by flow cytometry where cells were sorted based on size and fluorescence.

Flow cytometric cell counts of A. marina revealed that immobilized cells reached higher maximum cell densities when grown under NIR [(2.83 ± 0.005) × 10⁸ cells bead⁻¹ at 48 days, n = 3] (Fig. 3C) than under visible light [(1.46 ± 0.11) × 10⁸ cells bead⁻¹ at 47 days, n = 3]. The per-cell content of chlorophyll under visible light was 0.07 pg cell⁻¹ (Chl d) and 0.017 pg cell⁻¹ (Chl a). Higher Chl d concentrations were obtained for immobilized cells grown under NIR (0.2 pg cell⁻¹), while the Chl a per-cell content was found to be lower (0.007 pg cell⁻¹) (Table 1). Accordingly, Chl a/d ratios were >10-fold higher in immobilized cells grown under visible light (0.229) than under NIR (0.033). Based on the linear increase of cell numbers displayed in Fig. 3C, we calculated maximum cell-specific growth rates of 0.64 and 0.38 for NIR and visible light, respectively (Table 1). The corresponding population doubling times were 1.09 days (NIR) and 1.82 days (VIS).

Table 1.

Growth and photosynthesis parameters for alginate-embedded Acaryochloris marina cells^a

Parameter	NIR	VIS	Blue light
Maximum growth rate (μ)	0.64	0.38
Population doubling time (d)	1.09	1.82
Cells per bead	1.46 × 108 (day 47)	2.85 × 108 (day 48)
Chl a per cell (pg)	0.0070	0.0175
Chl d per cell (pg)	0.2077	0.0767
Chl a/d ratio	0.0339	0.2290
Pmax (μmol O2 mg Chl d−1 h−1)	1,272		1,128 ± 568

^aAll measurements are based on the average values of three technical replicates.

Wavelength-dependent O₂ evolution and photosystem saturation.

Oxygen profiling in NIR-irradiated beads demonstrated maximum O₂ production at a depth of ∼400 to 600 μm (Fig. 4). Gross photosynthesis (P) measurements, using either blue or NIR light, were performed by positioning the O₂ microelectrode tip at this depth and performing experimental light-dark shifts with the sensor remaining in place. P was normalized to the Chl d concentration within one bead (mg cm⁻³), yielding final units of μmol O₂ mg⁻¹ Chl d h⁻¹. Maximum gross photosynthesis (P_max) rates were on the order of ∼1,272 μmol O₂ mg Chl d⁻¹ h⁻¹ (n = 1) and ∼1,128 ± 568 (n = 3) μmol O₂ mg Chl d⁻¹ h⁻¹ for NIR and blue light, respectively (Table 1). No saturation of P was observed when beads were illuminated with the maximum NIR irradiance possible (365 μmol photons m⁻² s⁻¹) (Fig. 5A, NIR), whereas blue light saturated P at an irradiance of 144 μmol photons m⁻² s⁻¹ (Fig. 5A, blue light).

Fig 4.

(A) Cross section of an approximately 4-week-old alginate bead with clearly visible clusters of A. marina cells. (B) Oxygen concentration microprofiles measured with a Clark-type O₂ microsensor. Oxygen concentrations (in μmol liter⁻¹) were measured at increasing depths under six different NIR irradiance levels (13, 20, 25, 76, 200, and 365 μmol photons m⁻² s⁻¹). Microprofiles represent average values from three technical replicates.

Fig 5.

Photosynthesis of A. marina in alginate beads. (A) Gross photosynthesis was measured with an O₂ microsensor placed at a depth of 2 to 500 μm using the light-dark shift method. NIR was administered in increasing intensities by altering the distance of three collimated LEDs to the beads, while blue light was provided through an optical fiber placed directly at the bead surface. (B) Relative electron transport rates (rETR) measured on the bead surface using a fiber-optic PAM fluorometer. rETR was measured after a steady-state condition for O₂ production under either NIR or blue light was observed. Three independent biological replicates were used to generate the data displayed in panels A and B.

The onset of saturation, as determined through variable chlorophyll fluorescence measurement of rETR, was observed at a photon irradiance of ∼250 μmol photons m⁻² s⁻¹ of blue light (Fig. 5B, blue light), while no onset of saturation was observed for beads previously irradiated with NIR (Fig. 5B). Despite a lower photon irradiance, an ∼40% higher rETR was recorded when immobilized cells were irradiated with the maximum NIR irradiance (rETR, ∼162 ± 10, n = 3) than in those given blue light at a maximal irradiance (rETR, ∼115 ± 14, n = 3). Gross photosynthesis-versus-irradiance data (Fig. 5A) were used to calculate the relative photosynthetic efficiency (α) from the initial slope of the P versus E curves (Table 2). The α values were found to be higher in beads irradiated with NIR [(268 ± 0.29) × 10⁻⁶ m² mg Chl d⁻¹, n = 3] than in those irradiated with blue light [(231 ± 0.22) × 10⁻⁶ m² mg Chl d⁻¹, n = 3].

Table 2.

Comparison of relative photosynthetic efficiencies (α) of alginate-embedded and planktonic cells of Acaryochloris marina

Growth mode	α (m2 mg Chl d−1)a			nb	Reference or source
	Blue light	Red light/NIR	Visible light
Embedded	(231 ± 0.22) × 10−6	(268 ± 0.29) × 10−6		3	This study
Planktonic		438 × 10−6	298 × 10−6	4	18
Planktonic			205 ×10−6		42

^aAll measurements are based on three technical replicates, and deviation from the mean is shown for measurements performed in this study.

^bn, number of replicates.

DISCUSSION

The underside of didemnid ascidians is one of the few habitats of A. marina that has been described in terms of its light and O₂ microenvironment; here Acaryochloris grows in dense clusters interspersed with a diverse community of other microbes in a microniche enriched in NIR and depleted of visible light while experiencing irradiance-dependent shifts between anoxic and hyperoxic conditions (7, 22, 24). The same preference of A. marina for NIR-enriched environments was shown in an endolithic microenvironment (6), and a global distribution of Chl d-containing cyanobacteria in such niches was hypothesized (6, 20). Due to its unique photosynthetic apparatus, with Chl d dominating both PSII and PSI, the interest in A. marina has grown substantially, and laboratory-based studies have revealed many new insights into the limits of oxygenic photosynthesis. To our knowledge, all of these studies have been performed using planktonic cultures and have, as such, dismissed the naturally occurring growth state of A. marina. Our study is the first attempt at studying cells of A. marina under conditions mimicking their natural surface-associated growth mode, by embedding cells into an alginate matrix. We followed the growth of immobilized A. marina cells using imaging, photopigment extraction, and microsensor measurements.

Microscopy.

The growth of immobilized A. marina cells closely resembled the previously reported growth of alginate-immobilized bacteria and microalgae (55): immediately after inoculation, a few individual cells are dispersed throughout the matrix, and over time these cells grow into larger microcolonies that increase in size and cause expansion of the surrounding alginate. Eventually these microcolonies reach the outer surface of the bead, where they rupture it and release cells into the surrounding medium. This leakage is supported by alginate bead growth models (56) and by observations from immobilized microalgae (25). SEM of biofilms sampled from the underside of the didemnid ascidian Lissoclinum patella (7) revealed dense clusters of cells, approximately 1 to 2 μm in size, closely resembling previously reported sizes and micrographs of A. marina-like cells (6, 14, 32). Comparison with SEM images of the artificially immobilized A. marina cells revealed a high similarity, indicating that alginate embedment provides a good natural growth model for A. marina. Growth of other, unidentified bacteria was observed in beads imaged with SEM. This was expected, as most A. marina cultures are known to be nonaxenic (32, 48). Surprisingly, growth of these unidentified bacteria was only observed in beads kept under visible light and not in those kept under NIR. We did not further characterize these bacteria, which were relatively low in numbers, but their apparently wavelength-dependent growth suggests that they may be phototrophs unable to use NIR. Additionally, embedded A. marina cells grew to higher densities under NIR. Both of these aspects would enable A. marina to outcompete other members in the nonaxenic culture, and we speculate that repeated immobilization combined with incubation under NIR could thus be used for obtaining pure cultures of A. marina.

Photopigment analysis.

The blue and NIR light used in our experiments can be absorbed by chlorophylls (i.e., ∼440/675 nm for Chl a and ∼480/715 nm Chl d), enabling A. marina to use blue light, NIR, and other wavelengths of PAR for light harvesting and growth under different light sources. Two studies of A. marina reported doubling times in planktonic cultures under visible light ranging from 55 to 70 h (48) to 33 to 87 h (18), corresponding to 2.2 to 2.9 days and 1.3 to 3.6 days. These growth rates are comparable to the doubling times in biofilms obtained in this study (1.09 and 1.82 days for NIR and visible light, respectively). In terms of photopigment content, we found that alginate-embedded cells grown under NIR produced approximately twice as much Chl d as those exclusively grown under visible light. The Chl a concentrations measured in immobilized cells grown under visible light and NIR were remarkably similar (0.19 μg Chl a bead⁻¹ at day 40 for both), highlighting the importance of Chl d for both light harvesting and photosynthesis in A. marina. In A. marina, Chl a has essentially been displaced by Chl d as the main photosynthetic pigment (10, 11, 24, 34, 48–50) and its functional role appears more related to Chl d biosynthesis (45) and charge recombination in photosystem II (50). Photoacclimation in the A. marina type strain MBIC11017 appears to be performed through alterations of Chl a/d ratios under high-light (10, 32) and low-light (18, 48) conditions, while chromatic photoacclimation in A. marina was reported to involve regulation of its phycobiliprotein content (17). The Chl a/d ratios for immobilized cells grown under NIR were surprisingly similar (in our study, 0.0339) to previously published results (0.0346), where A. marina was grown under red light (18). Immobilized cells exposed to visible light (20 to 40 μmol photons m⁻² s⁻¹) had substantially higher Chl a/d ratios (0.229) than reported by Gloag et al. (0.0393 at ∼5 μmol photons m⁻² s⁻¹ and 0.0552 at ∼100 μmol photons m⁻² s⁻¹) (18) and were approximately twice as high as those reported in reference 32. Higher chlorophyll concentrations are commonly found in immobilized microalgae than in planktonic growth (25, 35). Our observed change in Chl a/d ratios could, however, also be related to self-shading effects of cell clusters throughout the beads (25), and higher chlorophyll concentrations do not necessarily relate to higher biomass but may rather represent acclimation to such self-shading within the beads.

By assuming an equal distribution of photopigment throughout the bead, we found that the per-cell Chl d concentration reached values of 0.2 pg Chl d cell⁻¹ under NIR and 0.07 pg Chl d cell⁻¹ when grown under visible light. A study on planktonic cells of A. marina (18) demonstrated comparable pigment levels for cells grown under visible light (0.03 to 0.05 pg Chl d cell⁻¹) but much lower values in cells exposed to red light (0.058 pg Chl d cell⁻¹). Gloag et al. (18) used tinted window glass for their growth experiments (30% transmission, >650 nm), and as Chl d absorbs maximally at 710 to 715 nm, higher Chl d yields were expected in immobilized cells grown under monochromatic NIR centered at 720 nm.

Photosynthetic activity and efficiency.

The initial linear slope of P versus irradiance curves, α, was used to assess the relative photosynthetic efficiency of immobilized A. marina cells. We found α values for immobilized A. marina cells irradiated with blue light that were lower than the α values determined under NIR illumination. Similar effects were shown for red light illumination in a study of planktonic A. marina (18). The α values for planktonic cells (298 × 10⁻⁶ m² mg Chl d⁻¹, grown under visible light) (18) were similar to the numbers obtained from another study using a waiting-in-line model (205 × 10⁻⁶ m² mg Chl d⁻¹) (42) and were overall comparable to the α values obtained for immobilized cells irradiated with blue light in our study [(231 ± 0.22) × 10⁻⁶ m² mg Chl d⁻¹]. The values of P_max for immobilized A. marina cells were calculated to be ∼1,128 μmol O₂ mg Chl d⁻¹ h⁻¹ for blue light and ∼1,272 μmol O₂ mg Chl d⁻¹ h⁻¹ for NIR. In another study, planktonic A. marina demonstrated much lower maximum gross photosynthesis values of ∼81 μmol O₂ mg Chl d⁻¹ h⁻¹ and 64 μmol O₂ mg Chl d⁻¹ h⁻¹ under visible and red light, respectively (18), but these studies did not match the illumination wavelengths to the absorption spectrum of A. marina as narrowly as in our study. P_max values of >200 μmol O₂ mg Chl d⁻¹ h⁻¹ were reported in reference 21, studying dense planktonic A. marina cultures in a microrespirometer. Diffusion limitations in the alginate matrix coupled with a high cell density of photosynthetically active A. marina within the alginate beads account for our finding of steep O₂ gradients similar to those of other photosynthetic biofilms (23). Apparently, the photosynthetic activity of A. marina under such conditions was not inhibited.

The use of pulse amplitude-modulated (PAM) fluorometry has been established as a valid tool to assess photosynthetic efficiencies in A. marina (18). Photosynthesis in immobilized cells of A. marina saturated earlier when previously exposed to blue light, while preceding NIR exposure caused a drastic increase in rETR values. No scalar irradiance measurements were performed within beads, and so we can only speculate on the light attenuation capacity of alginate. It is possible that NIR is less attenuated by alginate than blue light, thus penetrating more effectively and deeper into the alginate matrix. Deeper penetration of NIR than of PAR has been reported in coral skeletons (26) and didemnid ascidians (22, 24), which could explain our finding of earlier light saturation under blue light in embedded A. marina cells.

More accurate measurements of the efficiency of photosynthetic O₂ production in biofilms of A. marina have yet to be performed; such measurements are not trivial as they need, for example, to address the presence of light gradients and combined effects of light scattering and absorption. However, our results provide the first evidence for efficient NIR-driven photosynthesis in A. marina when growing in biofilms. It was recently shown that the energy storage efficiency of the photosynthetic light reactions in A. marina is comparable to or higher than that of typical Chl a-utilizing oxygenic phototrophs (29), suggesting that oxygenic photosynthesis is not fundamentally limited by the photon energies used by A. marina. Experimental approaches developed for use in microbial mats (3) showed a relatively low energy storage efficiency of different surface-associated microbial communities at the system level under high-light conditions (4). The same study also revealed higher photosynthetic efficiencies under low-light conditions. Similar studies of energy storage efficiency in A. marina biofilms have yet to be carried out.

We showed that alginate-embedded cells of A. marina grow well in a dynamic O₂ microenvironment experiencing hyperoxic conditions in the light and hypoxia in darkness. Preliminary microsensor studies of A. marina biofilms on the underside of Lissoclinum patella showed similar light-dark dynamics in O₂ concentration. Growth in alginate beads may thus be a suitable method for studying A. marina under conditions that mimic its natural growth mode and microenvironment in the laboratory under defined and reproducible conditions. This new way of growing A. marina also enables metabolic measurements in combination with gene expression studies under various experimental conditions. Such studies could give further insights into potential differential gene expression of A. marina in its planktonic and biofilm growth mode, which would provide further information on the niche adaptation and metabolic capabilities of A. marina in its natural habitat. It would, for example, be interesting to investigate and compare the microenvironmental regulation of the recently reported N₂ fixation in an A. marina isolate (36) under both planktonic and biofilm growth conditions.